Understanding health economics in clinical trials

As member of the Health Economics Analysis and Research Methods Team (HEART), together with my colleagues, on Tuesday 2 July I took part in a 1-day introductory short course entitled “Understanding health economics in clinical trials”, which was designed and delivered by the team. HEART is a new group of health economists who are based in UCL’s Institute of Clinical Trials and Methodology (ICTM), led by Rachael Hunter, and is involved at different levels in the economic components of clinical trials in different trial units at UCL. This short course was aimed at ICTM staff who are not health economists (e.g. trial managers, CIs/PIs, statisticians, data managers, research assistants, etc.) and was designed in response to the need we have identified over the last few years in working on trials as well as in response to colleagues across ICTM. This course was unique as it was intended specifically for non health economists working in trials, who wish to better understand the health economics in their study, and/or the health economist on their study. The course used a mix of lectures, group discussions and practical exercises to help participants consolidate their learning and see how to apply information from the sessions to real studies. No prior knowledge of health economics was assumed.

I believe the course was a success both in terms of the quality/quantity of the material covered during the six sessions throughout the day, as well as in terms of the positive feedback we received from the participants (almost entirely women, with the exception of two men). Many key and typically not well understood economic topics were discussed during the day, e.g. what are and how QALYs and costs are calculated, the potential limitations and issues of an economic analysis within a trial, or the role played by the protocol and analysis plan in the economic evaluation. My session was related to reporting and interpreting health economic results and I realised that most people who do not routinely deal with health economics may find difficult to grasp certain concepts or tools used in the economic analysis (e.g. what is a cost-effectiveness acceptability curve and how it can be computed). Nevertheless, I must admit that I was surprised by how many people were very motivated to learn these concepts and these “difficult” methods, often asking questions and making good comments (despite the fact that my session was the last of the course at the end of the day).
We ran this course as a trial as we did not have clear ideas of what an optimal design should be or the number of topics that should be covered for this type of course. We are now confident that the course has a solid structure and that there is a clear demand to learn the basic concepts of health economics, at least among people involved in trial analyses. Following the successful delivery of the course, we are planning to replicate the experience in the future, improving certain aspects of the sessions based on the feedback we received and also considering to open the course to meet the demand of a wider audience.

I have to say that this was an extremely positive experience for me as it was the first time I was involved in this type of projects. Me and my colleagues worked hard to design and prepare the different sessions of the course over the last few months, find the best way to link the arguments across the sessions, provide interesting group activities and materials for the practicals, etc. I have to thank all my colleagues who contributed to the promotion and realisation of this project, with a special mention for Caroline Clarke, who spent a lot of time and effort to organise the course and who personally contributed in giving one of the session of the course. Finally, I would also like to thank my colleague and health economist Ekaterina, with whom I had the pleasure to share the presentation and practical of my session in the course.

Perhaps the only true negative aspect of the course was the absence of a Bayesian perspective, especially related to the interpretation of the results and the statistical methods that can be used to perform the analysis. Given the generally low familiarity of the people attending the course with statistics, I believe it was reasonable not to further confuse them with another new element into the picture. However, I truly hope that people will become more and more familiar with the importance of using tailored statistical methods in economic evaluations to avoid biased results, and from that point to justify a Bayesian approach, well, at least for me, the step is straightforward!.